In vitro genebank study. Background and Justification

Transcription

1 In vitro genebank study. Background and Justification One of the significant outputs from the TaroGen project was the collection of 2,200 taro accessions from several Pacific Island countries. The strategy adopted by the project was to reduce this collection to a core collection of approximately 220 accessions, using morphological descriptors and DNA markers. Although this core collection is significantly less in number than the original collection, it is still relatively large and has to be conserved by as safe and secure a method as possible. As taro is a vegetatively propagated crop, it is traditionally conserved in field collections. This is not an effective conservation strategy for medium to long term preservation of taro genetic resources. Field collections are notorious for their demands on resources, with the safe maintenance of the accessions being reliant on constant vigilance and upkeep. Pest and disease outbreaks can be disastrous reducing the number of accessions in a collection to a level that is not representative of genetic diversity. Similarly, the climatic extremes experienced in the Pacific can also have a very significant negative impact on a field collection. More recently, as in the Solomon Islands, civil strife can be added to the list of events, which can seriously impact on field collections. As well as the need to identify an appropriate conservation system for this taro collection, the project also wanted to provide information to the countries on the conservation methodologies for taro, so that should they wish to maintain national collections they would have the information required for decision making. One key area of importance here is costs. There is very little information available on the cost of implementing any conservation strategy. Consequently, as part of the conservation component within the TaroGen project, a pilot in vitro genebank (IVAG) study was established at the SPC Regional Germplasm Centre (RGC) in parallel with a field genebank at Koronivia Research Station. The aim of the study was to determine all the inputs required for an in vitro genebank, and to compare this with the inputs necessary for a field genebank with the same accessions. In addition, there was also a need to evaluate the slow growth methodologies that had been developed for taro in another laboratory. Other issues that required consideration were genetic stability, safety and management. Taro is a crop that is amenable to tissue culture. Propagation systems have been developed that are optimal for a wide range of cultivars. In addition, it also seems to be a relatively stable crop, both in the field, and in tissue culture, so is suited to in vitro storage. A recent status report on the use of in vitro techniques for the conservation and use of plant genetic resources (Ashmore, 1997) concluded that slow growth techniques are now successfully and routinely applied to a range of species and across a range of genotypes within species. The use of these techniques was allowing storage of healthy germplasm with extended subculture intervals, thus reducing time and costs for maintenance. However, the report also identified gaps in both the development of techniques, as well as limitations in basic scientific knowledge. Some of the gaps highlighted are relevant to

2 this study, in that the proposed outputs would generate the information required to fill these gaps for taro. These were: More studies are needed on genetic stability after relatively long periods of storage, to establish the safety of slow growth storage, when compared with other methods. There is a need for the development and application of characterization systems, including molecular genetic markers for initial identification as well as monitoring of the genetic stability of stored accessions. Reproducible, simple and more widely applicable techniques are required. Cost analyses of in vitro conservation is needed to allow comparison with other conservation methods, particularly the comparison between in vitro storage and field genebanks. In vitro conservation methodologies offer an opportunity to preserve taro within a controlled environment, thereby eliminating risks from pest and disease outbreaks, and climatic extremes. Although various methods exist for reducing the growth rate of plants in vitro, the most widely applied slow growth storage technique is temperature reduction, often combined with a decrease in light intensity or culture in the dark There are reports in the literature of taro being stored for more than eight years at 9 C in total darkness, with transfer intervals of approximately three years. (Bessembinder et al., 1993). Similarly, Staritsky et al., (1986), reported that taro (Colocasia esculenta) could be conserved for three years at 9 C, and still be viable. Research carried out in the regional tissue culture laboratory at the University of the South Pacific, Samoa, when it was funded by the European Union, under the Pacific Regional Agricultural Programme, demonstrated that temperature reduction was the most practical method for slowing down the growth rate of taro. Taro could be maintained at 20 C, for 9 to 12 months, without subculturing, depending on the variety. Other parameters, besides reduced temperature, were also investigated. These were reduced light and supplementing the culture medium with osmoticums. The inclusion of mannitol in the culture medium did suppress growth, but some morphological changes in the resulting plantlets were observed. In addition, when mannitol was used with cultures initiated directly from the field, a phytotoxic effect was observed. As a result of the work carried out in Samoa, together with evidence from the literature, it was decided that temperature reduction would be used in this IVAG study. This pilot study was modelled to some extent on a study carried out by CIAT in collaboration with the then IBPGR over a three year period. (IBPGR/CIAT, 1994). The study covered the principal steps involved in an in vitro active genebank, from the selection and sampling of material from the field to the evaluation of genetic stability of retrieved material growing under slow growth conditions. The overall objective was to assess and demonstrate the technical and logistical aspects of establishing and operating a pilot in vitro active genebank using cassava as a model crop. The pilot genebank used 100 clones selected from the 4,250 cassava accessions maintained in the field. At the end of the study several recommendations were made regarding in vitro conservation by slow growth. These were: The value of this approach should be balanced against other conservation strategies, based on knowledge of the genepool, reproductive behaviour, range of variability

3 (geographic and genetic), and costs. It should complement other conservation strategies for the same crop species, such as seed, field genebank, and in situ conservation. There should be a thorough knowledge of the in vitro culture behaviour, such as culture initiation, explant, micropropagation, and requirements of the species, prior to any in vitro conservation strategy. Depending on the crop and its agronomic and economic importance, two different levels of in vitro conservation should be envisaged: a fully implemented system compared with a minimum input system. All the steps, procedures, and data required throughout the experiment should be taken into account. The rate at which accessions are introduced into the in vitro genebank should be balanced against the risks of losing accessions from genetic erosion, pest and disease attack or climatic extremes in the field. It is important to introduce clean material into the in vitro genebank despite the delays this might impose on the study. Genetic stability is an important consideration of any in vitro conservation strategy. Monitoring techniques for stability will depend on the history of the crop species regarding this trait. Sophisticated, high-technology monitoring has value, when there is sufficient base to assume instability, either intrinsic, or due to the culture system and subculture frequency. In other instances, visual observation of morphological changes may be sufficient in comparison with appropriate controls. If variants are observed in the in vitro system, the reason for the variation has to be determined, and in this case the best option is to refer back to the original field collection. A field collection should exist for as long as the in vitro genebank has not been duplicated elsewhere for security reasons. Once a collection has been duplicated, only material for evaluation should be held in the field. Decisions regarding the number of replicates will depend on: size of the collection, size of vessels, risk of losses during in vitro multiplication, subculturing and storage. In the case of cassava, a minimum of one and a maximum of three replicates were lost, and therefore a replication of three to five per accession is recommended. The information system required for the management of the collection will depend very much on the size of the collection. These recommendations were taken into consideration when the pilot genebank study for taro was established under the TaroGen project. Objectives of the TaroGen pilot in vitro genebank (IVAG) study The overall objective of this project is to investigate the technical and logistical aspects of establishing and running an IVAG, using taro as a model. Specific To select a sample of Fijian taro accessions, and to process these samples into in vitro storage under conditions of slow growth.

4 To provide a cost analysis for both field and in vitro maintenance of taro To evaluate the applicability and reliability of available slow growth methodologies for taro. To monitor genetic stability (through visual observation), and viability during slow growth in in vitro storage. To determine the needs for laboratory facilities, equipment, consumable items, and technical staffing involved throughout the operation of the in vitro genebank. To provide recommendations for establishing and running an in vitro active genebank (IVAG) on the basis of the experience gained with taro. In vitro genebank Selection and multiplication of accessions 50 of the non-export taro accessions were initially selected from the field collection at Koroniva Research Station. However, because of delays in obtaining sufficient suckers for initiation into tissue culture, only 44 accessions were eventually established in tissue culture. Shoot-tips were excised from these suckers and subjected to a three stage sterilization procedure, consisting of 70% alcohol for 2 minutes, followed by 20% domestic bleach (1% active chlorine) for 15 minutes, and after trimming, another wash with 10% domestic bleach (0.5% active chlorine) for 10 minutes. The final stage was 3-4 washes with sterile distilled water. From previous experience with in vitro culture of taro, and the information from the CIAT study on cassava, it was decided that for every accession in the pilot study there should be five replicates. Bulking up was achieved by using the three-stage multiplication system, which utilizes the growth regulators thidiazuron (TDZ) and benzylaminopurine (BAP). This involves culture on a medium containing TDZ for three to four weeks, followed by culture on a medium supplemented with BAP for a further three to four weeks. For the next stage the explants are transferred to a medium containing a low concentration of TDZ. Depending on the degree of multiplication achieved, the cycle can begin again after the third stage, or if the multiplication rates are sufficient, the explants can be cultured on a medium containing the basic macro- and micro-nutrients, and no growth regulators (Murashige and Skoog, 1962) Multiplication rates were affected by the presence of endogenous bacteria in the explants. These bacteria are in the plant tissue and are not eliminated as part of the surface sterilization procedure. Tissue cultures of taro were analyzed by CABI, UK in mid-2000, and the bacteria identified as Pseudomonas, Bacillus, Micrococcus, Curtobacterium and an unknown coryneform. Some taro cultures had been sent to the UK in the early 90 s for bacteria identification, and at this time the bacteria present were identified as Pseudomonas and Methylobacterium spp (D. Stead, pers. comm). In the IVAG. these bacterial contaminants did not express themselves immediately, and so the problem was not apparent until after several subcultures, and some varieties were affected more than others. Because of the need to generate numbers for the IVAG, the possibility of eliminating the bacteria through meristem culture could not be determined. For this

5 reason, antibiotic treatment only was investigated. Rifampicin was the first antibiotic investigated, as it had proved effective in control of bacteria in sweet potato cultures at CIP and also in the USP laboratory in Samoa. Treating contaminated taro cultures with 80mg/l rifampicin eliminated the contamination from 89% of the 65 explants tested. However, after culturing with 0.5mg/l thidiazuron for multiplication, contamination reappeared in the majority of the explants. The effectiveness of gentamycin was also evaluated. This was chosen as it had been shown to be effective with cultures of other plant species, where Pseudomonas spp had been shown to be present. Using gentamycin with rifampicin, it was possible to eliminate the bacteria from the cultures and proceed with the multiplication process. Slow growth storage system The RGC has one growth room, which runs at a temperature of 20 C, and so establishing an IVAG at a reduced temperature presented no problems. Shoot-tips of approximately 1cm in size were excised from the cultures that had been on the multiplication cycle. Prior to their excision all cultures after the multiplication period, were grown on basal medium without any growth regulators for one month, to reduce any possible carry-over of the growth regulators into the IVAG. In the IVAG taro cultures were grown in glass jars of volume containing 20mls of Murashige and Skoog basal medium (1962), supplemented with 3% sucrose. This medium was changed after three months of culture as it was apparent that growth was too vigorous, especially root production. The IVAG was re-established using the same basal medium but supplemented with benzylaminopurine (BAP) and napthaleneacetic acid (NAA). Evaluation of viability Some guidelines were necessary for when the security/survival of an accession was under threat. It was therefore decided that cultures would be replaced when three or more cultures no longer looked healthy, that is, were not viable. Viability here refers to cultures that had either (a) not grown from initiation, or (b) had outgrown the culture container and the culture medium, and were starting to senesce, or (c) were affected by any of the viability factors below. This was a parameter set in an IPGRI-coordinated sweet potato slow-growth experiment. Once three or more of the replicates are no longer viable the security of that accessions is affected. Viability descriptors were recorded every month. These were: shoot tip necrosis: this would be observed as a form of dieback from the main shoot, and would be recorded as absent (0) or present (1). stunting: stunting is a concern for the reasons outlined above. Cultures that are too stunted would not grow into plants. contamination: if contamination was present it was recorded as 1, if absent, then 0 would be recorded. A separate sheet was kept for noting what the contaminant was (bacterial or fungal), and what the treatment was. senescence: this was recorded as number of leaves that had senesced out of the total number of leaves. These non-viable cultures would be removed, and that accession would be replaced with five new replicates, either generated from the cultures in the IVAG, or from the

6 cultures maintained outside of the IVAG. Contaminated cultures similarly were removed from the growth room. If the problem was fungal contamination, then the culture would be cleaned (provided fungal growth was not excessive), and returned to the IVAG. All inputs required for that cleaning process were recorded. Any cultures with bacterial contamination were destroyed and replaced with new cultures from the multiplication phase. Morphological characterization in vitro. With the IBPGR/CIAT study on cassava, morphological characters were observed after six months of storage. The parameters examined were selected as being relevant for cassava, and included etiolation, shoot number, callus formation, rooting, aerial roots, and special observations such as different leaf shape, pigmentation etc. Different parameters were chosen for taro. These were: sucker number: the number of suckers per replicate was recorded. callus formation: although callus is uncommon in shoot-tip cultures of taro, it was included because of its importance in terms of genetic stability. The values used were 0, indicating absence, and 1 for presence. rooting: rooting was evaluated in terms of relative amount, 1, 2, 3, which corresponded to poor, medium and high respectively. hyperhydricity: this was included as it can be a problem with cultures that have had exposure to TDZ, and can affect the establishment of that culture in the soil. stunting: this was a problem experienced with an exotic variety in Samoa, in that the cultures could not be induced to develop, and elongate into viable plantlets. leaf shape: leaf shape can be an indication of some form of change. When experiments were carried out using mannitol with taro, leaf shape was affected. These parameters were monitored after three months of culture Genetic stability It was hoped that the cultures subjected to slow growth storage could be analysed using the DNA fingerprinting system developed for the taro collections. However, this was not possible, and so stability was monitored through observation of morphological characters. It is also intended that a random sample will be planted out in the field, and morphological descriptors recorded for these, and compared with those plants that have experienced no in vitro processing and have been continuously in the field situation. It is worth noting here that in one of the recommendations from the IBPGR/CIAT report, it was commented that the use of DNA technology is only really relevant if there is sufficient basis to suspect instability, otherwise visual observation of morphological changes in comparison with appropriate controls is adequate. Taro is a crop that appears to be relatively stable in the field, and has not demonstrated loss of genetic integrity in tissue culture. With some work in Samoa a change in petiole colour was observed with one of the varieties in tissue culture. This variety was monitored and characteristics such as yield weight, shape and taste recorded. These all gave the same results as those with normal petiole colour.

7 Associated field genebanks In the IBPGR/CIAT study, Associated Field Genebanks (AFG) were established. These consisted of (a) 100 clones without any in vitro processing, (b) the same 100 clones but subjected to a micropropagation stage and (c) accessions after the in vitro storage period. Although the rationale behind the need for AFGs is sound, one of the comments from the report was that the establishment and running of the AFG is expensive and time consuming and that its function can be replaced by the original field collection. Considering the limited resources of the project, and the even more limited resources available at Koronivia Research Station, the only AFG established was the accessions without any in vitro processing. Costs of in vitro culture Although it is generally agreed that in vitro storage is a more secure method of conservation, it is often argued that it is too expensive to consider, especially in regions and/or countries where resources are limited. However, there is very little evidence to support this argument, therefore there is a need to provide more cost analysis information for the various conservation strategies that are in use. In addition, as the number of accessions increase, and organizations/institutes make a commitment to the conservation of countries germplasm, cost analysis becomes more crucial. A cost analysis identifies areas where cost controls and reductions can be implemented. Such analyses suggest ways to achieve cost effectiveness. Perhaps certain consummables can be changed so that costs are reduced; until an accurate cost analysis is carried out, unnecessary expenditures might slip by unnoticed. Cost analysis also assists with charging for tissue cultures, if some aspect of the tissue culture laboratory is to be commercialised. A standard cost analysis encompasses total, variable and fixed costs. Variable costs involve inputs that are easily varied in the short term, usually a period of less than one year. Fixed costs involve items that cannot be varied in the short term, for example, buildings and machinery. Straight-line depreciation can be used to determine annual depreciation expenses for capital goods such as buildings, machinery and equipment. This type of cost analysis was carried out by CIAT for both in vitro and field collections (Epperson, et al.,1997). The total number of accessions for both collections was 5,992. The total cost per accession in the field was US$17.09 with variable costs running at US$ With the in vitro collection the total costs/accession were US$26.22 with variable costs running at US$1.85. Variable costs usually refer to the costs that will change depending on the number of accessions you have. The costs of the pilot genebank were assessed in two different ways: Using spreadsheets devised by the ACIAR funded project, Economics of preserving genetic diversity in PNG specifically for the cost analysis of the SPC RGC. These sheets cover all costs for in vitro conservation and distribution of taro at SPC, and consist of (a) a price list for variable costs, (b) a summary table, (c) key variables table, (d) initiation and maintenance budget, and (e) multiplication and distribution budget. The cost budgets contain data and formulae developed to estimate the total

8 cost of inputs employed in the conservation of in vitro collections, and the multiplication and distribution of the germplasm material. The summary table provides an annual estimate of variable costs, medium term variable costs, fixed costs and total costs for maintaining the in vitro taro collection at SPC. This summary table, which is included in this report, allows users to examine the cost budget without having to view the entire spreadsheet. Costs are estimated for the whole taro collection, per accession, and per plant replicate. A simpler system was devised so that all inputs in the IVAG were recorded at the time of carrying out the activity. This shows what resources (labour, equipment, consummables) are required for the different operations, and also confirmed the cost analysis from the spreadsheets. It also gives an indication of immediate costs. This costing does not take into account fixed costs. Results and Discussion: In vitro genebank Availability of material for genebank The availability of suckers for initiation into the multiplication cycle was a problem. With varieties that easily produce suckers there was sufficient material for introduction into the IVAG. However, some of the varieties chosen for the genebank were poor at suckering, and so it was difficult to obtain an adequate number of suckers for the experiment. It was also found that some of the selected accessions would not respond well to the multiplication system and so could not be used in the genebank. This reduced the number of actual accessions used to 44, instead of the original 50 selected. This same problem was experienced by CIAT in their pilot genebank experiment, of the 100 accessions selected for the study, three would not respond to tissue culture, and so could not be used. Contamination also affected the multiplication process and rates of the selected accessions. As noted earlier, this was contamination due to endogenous bacteria. Although for the purpose of the multiplication cycle, these bacteria were eliminated using the combination of the antibiotics, rifampicin and gentamycin, bacterial contamination caused a problem later in the study. Contamination occurred with both these cleaned-up cultures and also with other cultures where bacterial contamination had not been a problem in the initial stages of the study. It would seem therefore that in the cases where antibiotics had been used, they had merely suppressed the expression of the bacteria. In the plants where contamination was not visible initially, the growth regulators used in the multiplication process, and the rate of subculturing, encouraged the expression of the bacteria. Using antibiotics is not a viable option for cultures that are going to be used in a slow growth storage system. Evidence shows that generally the antibiotics merely suppress bacterial activity, and can lead to resistance problems at a later stage. In addition, there is the possibility that the use of antibiotics in combination with a system that induces stress through slow growth will result in genetic instability. Antibiotics are also costly, and so add an extra cost to the total in vitro conservation cost.

9 The recommended strategy for addressing endogenous contamination in cultures is to screen the material before it is initiated into tissue culture. A microbial detection medium is easy to make and this can be used to determine whether or not a sucker from a particular accession is contaminated. Any plants testing positive can be rejected and only those that test negative can be introduced into tissue culture. Initiating clean material into tissue culture will guarantee that multiplication rates are not affected, and that there will be no bacterial contamination problems with those cultures at a later date. This need to screen obviously has implications for plant numbers and again means that there has to be an availability of planting material. This planting material should also be easy to access so that there is a minimum delay between harvesting and the tissue culture process. Any extended delay will only exacerbate any bacterial contamination problems. Response to slow growth storage system When an in vitro genebank is established, one technique has to be selected and applied to all the different genotypes under investigation. Varietal influence can be responsible for very different responses to the selected slow growth methodology. CIAT found that of 48 varieties, 50% had to be subcultured after one year of storage, about 18 after 15 months, and 6 required subculturing after 8-9 months. In this study all of the varieties would grow without requiring subculture for six months, but after that 70% required subculturing whereas the remainder could be maintained in culture for a further three months. The need to subculture reflects to a large extent the vigour of the plant, in that the culture medium has been fully utilized, and nutrient deficiency is therefore a potential problem. In addition, with older cultures, defoliation and senescence become a more common occurrence, increasing the chance of fungal contamination from rotting leaves in the culture vessel. This lack of uniformity across the range of genotypes being investigated merely adds to the labour input required in successfully maintaining the genebank. Culture viability has to be checked regularly, and decisions made as to when subculturing is required. If this is necessary with individual accessions, or small groups of accessions, too often, then this will place a strain on resources. The system that was initially used with the taro from Fiji in this study was the same system used in Samoa with the regional taro collection held in the laboratory at the University of the South Pacific. Using the system there, the taro could be maintained without subculture for 9-12 months. With the Fijian taro, the need to subculture had been reduced to an average of six months instead of nine months. This is possibly due to the length of time the accessions in the USP laboratory have been in tissue culture, and so a reflection of their reduced vigour. The Fijian accessions, on the other hand, had been initiated into the IVAG more or less directly from the field. This shows that one cannot assume that a system that works in one laboratory, will work in the same way in another laboratory. Some of the laboratories in the region carried out an experiment on slow growth storage of sweet potato in collaboration with IPGRI, and the results from each laboratory where the same varieties were used with the same methodology, were quite different.

10 The inherent vigour in the explants resulted in excessive growth from the majority of the accessions, especially when the medium used was basal Murashige and Skoog. The IVAG was re-initiated with the same medium but supplemented with growth regulators, BAP and NAA. This did reduce the growth compared to the initial medium used, but as stated, the interval between subcultures was shorter than that experienced in the laboratory in Samoa. As a result of this different response to the culture medium used in the USP laboratory in Samoa, some smaller experiments were established in the RGC looking at other parameters for slowing down growth. These include culture size, reduced nitrogen in the medium and sucrose concentration in the medium. Contamination After each subculture, or re-initiation of the IVAG, fungal contamination was recorded within one to two months. The rate of fungal contamination was 4%, which is relatively high. However, once these cultures were removed, and replaced with new cultures, fungal contamination did not re-occur. It is likely that this fungal contamination was due to using old cultures, where fungal spores were present, but not visible, as the source of the explant. Bacterial contamination was less obvious, but was detected in approximately 8% of the cultures. However, when all cultures were screened for the presence of bacteria to determine whether reduced growth was due to genotype or the presence of bacteria, 20% were tested positive. CIAT found that contamination largely occurred in 7 th month of storage, and was much higher in some varieties than others. A problem that can occur with maintenance of cultures for a long period is defoliation and senescence, then cultures can become contaminated. This might not be obvious until the cultures are subcultured for reintroducing back into the IVAG. Therefore, there is a need to identify when a balance is achieved between having to subculture because of contamination risks, and efficient use of resources. Senescence occurred fairly early on in the growth of the cultures (after 2 to 3 months), but remained at 30-50% throughout the culture period. Cultures that suffered from fungal contamination were either cleaned up, or they were replaced with new cultures. All inputs required for the cleaning process were recorded. Any cultures with bacterial contamination were removed, and replaced with other clean cultures. Genetic stability It was intended that a sample from the IVAG could be sent for DNA fingerprinting, and the resulting fingerprints compared with fingerprints from the same accessions that had not been through the genebank. However, this was not possible. The cultures were monitored for selected characteristics for the duration of the experiment. At no time, was there any indication from these recordings that genetic stability was a problem. A random sample of accessions from the genebank has been retrieved from slow growth storage, transplanted to screenhouse, and will be planted in the field for morphological evaluation.

11 Costs These summary tables have been extracted from the ACIAR spreadsheets, and provide an annual estimate of variable costs, medium term variable costs, fixed costs and total costs for maintaining the in vitro taro core collection at SPC, and also the IVAG. Summary tables allow users to examine the cost budget without having to view the entire spreadsheet. Costs are estimated for the whole taro collection, per accession, and per plant replicate. Using this system the cost to maintain one accession for one year has been calculated at F$98.3. (US$49), if the RGC is maintaining the entire taro core collection of 206 accessions. (Table 1). This is a relatively high cost but as can be seen by a closer study of the costs involved, it is the fixed costs that are the major contributor to this figure. These fixed costs are very extensive, and include the original cost of converting the barracks to the buildings now used by SPC, and also the rent that SPC has to pay for the present location of the library, which used to be housed in this building. This is what is known as an opportunity cost. In addition, costs are also included for support given to the RGC by administration, finance and information technology, and these are quite high. These costs were taken from a 1999 budget for these departments, and then divided according to the number of staff at SPC. The actual tissue culture costs are relatively low. The cost of in vitro storage quoted by CIAT is slightly lower than this at US$26.22 per cassava accession, and this is probably due to CIAT being a much larger scale operation, and so the lower cost reflects economies of scale. Reducing the actual tissue culture costs for maintaining one accession would be quite difficult as they are already relatively low. Since this costing was first done, (approximately 12 months ago) the RGC has taken action to reduce costs by changing some of the consummables that are being used. Further attempts to reduce the cost would have impact on the high standards being practiced in the RGC. The impact of fixed costs is still seen when comparing Tables (1) and (2), where the number of accessions differ between the core collection of 206 and the IVAG of 44 accessions. Fixed costs with a collection of 44 accessions are still high. Cost per accession, when 206 accessions are in storage is F$98.3, (US$49), whereas when 44 accessions are in storage the cost is F$146.9 (US$73.08) per accession. This does show the effect of economies of scale. Once a laboratory is established, it is important to put it to full use to make best use of resources.

12 Table 1: Summary Table showing all costs for maintaining the taro core collection of 206 accessions at the SPC RGC Costs Cost of Dalo Collection/Year ($FJ) Cost per Accession/Year ($FJ) Cost per Plant Replicate/Year ($FJ) Variable Costs (A) Pathogen Testing (B) Initial Medium Preparation (C) Cleaning of Plant Material (i) Isolation from the field (ii) Ongoing cleaning of material (D) Preparation of Materials for Sub- 3, Culture (E) Laboratory Hygiene (F) Multiplication (G) Distribution Medium Term Variable Costs Total Variable Cost 3, (A) Cleaning of Material (B) Laboratory Hygiene (C) Stationery Fixed Costs Total Medium Variable Cost (A) Capital Costs (i) Laboratory equipment 1, (ii) Office equipment 1, (iii) Miscellaneous 8, (B) Labour Costs 5, Total Fixed Cost 17, Total Cost 20,

13 Table 2: Summary Table showing all costs for maintaining the taro core collection of 44 accessions at the SPC RGC Costs Cost of Dalo Collection/Year ($FJ) Cost per Accession/Year ($FJ) Cost per Plant Replicate/Year ($FJ) Variable Costs (A) Pathogen Testing (B) Initial Medium Preparation (C) Cleaning of Plant Material (i) Isolation from the field (ii) Ongoing cleaning of material (D) Preparation of Materials for Sub Culture (E) Laboratory Hygiene (F) Multiplication (G) Distribution Total Variable Cost 1, Medium Term Variable Costs (A) Cleaning of Material (B) Laboratory Hygiene (C) Stationery Total Medium Variable Cost Fixed Costs (A) Capital Costs (i) Laboratory equipment (ii) Office equipment (iii) Miscellaneous 2, (B) Labour Costs 2, Total Fixed Cost 5, Total Cost 6,

14 Table 3: Simple cost analysis (F$) for maintaining 20 taro accessions (5 replicates each) in slow growth storage (6 month subculture interval). Acc Nos Activity Time Labour pa Cons Cost pa Overall cost pa 20 Plant 3h20m preparation 20 Shoot-tip 3h20m excision 20 Media 3h preparation 20 Subculturing 5h Cleaning up 1h Contaminated plants 20 Monitoring 5h the genebank Initiation maintenance Maintenance These costs were obtained from the pilot genebank experiment and so reflect the amount of time that was involved in monitoring the experiment. With a reliable slow growth storage system in place, it would not be necessary to monitor the genebank so closely. It would just have to be checked regularly for contamination and viability. Subculturing costs are high because of the cost of the culture containers. Cost could be reduced here if old jars were obtained free of charge. This table only allows for variable costs, it excludes all medium term variable costs and fixed costs. There is also no allowance for electricity. This would vary significantly with each laboratory.

15 Resources Used Table 4: Activity table to show resources used for maintaining one taro accession (5 replicates each) in slow growth storage (6 month subculture interval) Activity Plant preparation Excision of shoot-tip and initiation of culture (includes sterilization) Preparation of 1l of medium Subculturing Cleaning up contaminated plants for 1 plant only (unlikely that entire accession requires cleaning) Recording Time required for 1 accession (5 replicates) 10mins 10mins (based on 4 accessions taking 40mins) 9mins 15mins 15min 1 5min The first two activities would be required every time a plant was initiated into tissue culture. Once the collection had been established, only the last three activities are required. Any replacement of plants in the in vitro genebank by plants from the field would require the first two activities again. Chemicals With the exception of the antibiotics required at the beginning of this study, the only chemicals required were those used in preparation of the culture medium. These include all the basic macro- and micro-nutrients as defined by Murashige and Skoog (1962), supplemented with sucrose and a gelling agent. For the multiplication phase of the study, Stage 1 and Stage 3 media were supplemented with thidiazuron (TDZ), whereas Stage 2 medium contained the growth regulator, benzylaminopurine. All of these chemicals, with the possible exception of TDZ would be found in any basic tissue culture laboratory. TDZ would be present if the laboratory was involved in active multiplication of taro, such as the lab at Nu u in Samoa. For the maintenance medium used in the IVAG, the basal medium was supplemented with benzylaminopurine and napthaleneacetic acid. Glassware The glassware used in the IVAG mainly comprised beakers, graduated cylinders, pipettes, conical flasks, glass jars (culture containers) etc., These would all be available in any basic tissue culture laboratory. Equipment Basic tissue culture equipment was required for this study, such as, electronic balance, ph meter, hotplate/magnetic stirrer, autoclave (pressure cooker can suffice) distilled water unit, LAF cabinet etc. Any reasonably equipped tissue culture laboratory would contain this equipment.

16 Facilities The following SPC RGC facilities were used in the IVAG: Tissue culture preparation room. Transfer room Growth room at 20 C. Screenhouse The field collections maintained at Koronivia Research Station were also used for the supply of the taro accessions used in the IVAG. Staffing. One technician had the responsibility of the IVAG. Because of other duties, the time allocation was not 100%. On average, the time spent on this study was 2%, but the technician had to be able to make larger amounts of time available when it was required. Obviously, the establishment of the genebank required a large time input and whenever the accessions had to be subcultured. The TaroGen Tissue Culture specialist supervised the study. Less supervision would be required if there had not been problems with contamination and plant vigour. If a system is established with all the necessary recommendations and all the parameters calculated minimum supervision would be necessary. Field genebank There is a more detailed paper on the field genebank, but the political situation in Fiji last year did emphasize the problems that can occur with field genebanks, if they cannot be given the attention they need. The genebank went through two cycles, but with both cycles there were problems. Some of the problems encountered were largely due to the political situation. These were: Reduction of budget allocated; hence, No casual labourers Untimely arrival of materials such as fertilizers and chemicals. High incidence of theft Trampling by cattle Increase in labourers absenteeism Inefficient control and management of labourers Extended wet weather. The in vitro genebank also suffered at this time due to lack of observation and monitoring. There were several days when staff had to stay at home because of security reasons, and when at work it was difficult to focus on work with the day s events. However, because the in vitro genebank is a far more controlled and self-sustaining environment, these problems had minimum impact, as is indicated by no lost accessions from the in vitro genebank, compared to four lost accessions from the field genebank.

17 Table 5: Summary Table showing all costs for maintaining 50 accessions in the field at KRS Activity Cost of Collection/ Year (F$) Cost of Each Accession/Yea r (F$) Cost of Each Plant/Year (F$) Variable costs [A] Land Preparation - Slashing (Machinery) Ploughing (Machinery) Harrowing (Machinery) Field Marking Total Cost of Land Preparation [B] Planting and Replanting - Planting Plant Rescue a In-filling Total Cost of Planting and Replanting [C] Crop Maintenance - Weeding Fertiliser Insecticides Fungicides Total Cost of Crop Maintenance [D] Harvesting Total Variable Costs Medium Term Variable Costs [A] Equipment [B] Stationery Total Cost of Medium Term Variables Fixed Costs [A] Capital [B] Labour Total Fixed Costs Total Costs This table shows all the costs for maintaining the 50 accessions at KRS for one year. As can be seen when comparing the costs here with the costs as shown in Table 2, there is very little difference between field and in vitro costs for a collection of 40 to 50 accessions. Again the impact of fixed costs are seen. The labour input in the fixed costs for the field genebank is fairly high and it is possible that reductions could be made here.

18 Assuming these costs are correct and labour input has been calculated accurately then, it does indicate quite clearly that the costs of the two forms of conservation are very similar. Depending on the circumstances, in vitro conservation could be a more sustainable use of resources because you have the added advantages of in vitro storage. Conclusion During the period of the IVAG, important components of the establishment and operation of such a genebank were assessed. These included: sampling of material from the field for the genebank, micropropagation of the accessions for storage, the applicability of the storage system, and the cost of maintaining an IVAG. Because of the problems generating sufficient material for storage, the IVAG only ran for 16 months. As a result, components of the IVAG, such as genotypic stability were not fully evaluated. Logistical aspects of in vitro storage like equipment needs, supplies and technical staff requirements were also determined in the period of the IVAG. Using the methodology evaluated in this study of low growth temperature and a basal medium supplemented with the growth regulators, BAP and NAA, the subculture frequency ranged from six to nine months; this variability was attributed to genotypic effects. No accessions were lost from the IVAG, but because of the genotypic effect and also the problem with bacterial contamination, replicates had to be replaced from outside the IVAG, that is, from the multiplication phase attached to the IVAG. Very little suckering was observed in any of the cultures. As the sample of accessions in the IVAG covered a range of different varieties with differing suckering ability, this was obviously due to the conditions of the IVAG, and was not the result of genotype. A preferable system could be one that encourages some suckering, yet at the same time, reduces apical dominance and rooting. In this way, the IVAG could always be supplied by material from the genebank, should replicates be lost. There would be less concern too, about old cultures becoming sources of contamination for the next culture. Morphological characterization relied on shoot number, callus formation, rooting, hyperhydricity, stunting and leaf shape. No changes were observed in these characters to indicate a problem with genetic stability. Actual maintenance costs for both systems are very similar, and the spreadsheets prepared for both types of genebanks enable a full analysis of all costs. As stated with the in vitro genebank it would be very difficult to reduce the actual tissue culture costs further. The major cost is with the fixed costs. This is the same with the field genebank, and so if reductions in costs are to be made with either method, it must be in the fixed costs (labour?). It could be argued that where conservation is concerned, in vitro is better value, as the accessions are protected from any pest and disease outbreak, or climatic disaster, unlike with the field genebank.

19 Recommendations The reason why in vitro storage is being used has to be clearly defined. The value of this approach should be balanced against other conservation strategies. In vitro storage enables germplasm to be maintained in a pathogen-tested state, which could provide farmers with a yield advantage over several years. This might also be important in larger island countries where certain diseases are found in only some areas. It also offers advantages on space and vulnerability to pest and disease attack, and climatic extremes. Any slow growth storage system will not work for all varieties to the same extent. There has to be allowances for this and this will impact on resources with the need to subculture different accessions at different times. Only those accessions that respond well to the existing multiplication system, and also the slow growth storage system should be put into in vitro storage. Using accessions with low responses to both will jeopardize the security of that accession. Accessions should be screened for bacterial contamination prior to being introduced into in vitro storage. This will eliminate any problems with bacterial contamination later. There should be relatively easy access to the material required for the genebank, as this will facilitate the sampling and screening process. A decision has to be made as to when is the optimum time to subculture. If cultures are left for too long, then this increases the risk of fungal spores being present to contaminate the next culture. A balance has to be achieved between this risk and use of resources, as the more subculturing required, the more resources are used. Five replicates would be the minimum number of replicates recommended for an in vitro genebank of taro. However, if there is sufficient material available for initiation into the genebank, then using more replicates would increase the security of that accession and lengthen the time period before subculturing is required. A complete field collection is necessary, at least until there is enough confidence in the reliability and stability of the in vitro system, or the in vitro collection has been duplicated for safety reasons. Whichever genebank system is selected, the manager/curator must be aware of all the costs and carry out a full cost analysis to ensure that the resources available will sustain all activities. The documentation system for managing the process depends to a large extent on the size of the collection and the resources available. It can be simple cards or a computerized system.

20 Proper labelling of plants throughout the system is necessary. Accessions should be given accession numbers that are specific to that genebank at the time of introduction to the genebank. Ideally these accession numbers should include letters and numbers, as this helps to prevent mis-labelling problems. The use of variety names should be avoided, as this can lead to errors. Close monitoring and accurate recording is required. References Ashmore, Sarah E Status report on the development and application of in vitro techniques for the conservation and use of plant genetic resources. International Plant Genetic Resources Institute, Italy. Bessembinder, J.J.E., G. Staritsky and E.A. Zandvoort Long-term in vitro storage of Colocasia esculenta under minimal conditions. Plant Cell, Tissue and Organ Culture. 33: Engelmann, F In vitro conservation methods. Pp In Biotechnology and Plant Genetic Resources: Conservation and Use. Biotechnology in Agriculture Series: 19 (J.A. Callow, B.V. Ford-Lloyd and H.J. Newbury, eds) CAB International, UK. Epperson, J. E. Pachico, and D. H. Guevara, C. L A cost analysis of maintaining cassava plant genetic resources. Crop Science. 37: 5, IBPGR/CIAT Establishment and operation of a pilot in vitro active genebank. Report of a CIAT-IBPGR Collaborative project using cassava, Manihot esculenta Crantz, as a model. IPGRI, Rome/CIAT, Cali, Colombia. Murashige, T. and F. Skoog A revised medium for rapid growth and bioassays with tobacco tissue cultures. Physiol. Plant. 15: Staritsky G., A.J. Dekkers, N. P. Louwaars and E. A. Zandvoort In vitro conservation of aroid germplasm at reduced temperatures and under osmotic stress. Pp In Plant Tissue Culture and Its Agricultural Applications. (L.A. Withers and P.G. Alderson eds) Buttersworth, London, UK